Judge’s sentencing order raises concern among prosecutors, defense lawyers

Listen to this article

Prosecutors and defense attorneys alike say they are concerned that a Superior Court judge’s recent order in an assault and battery case could lead to judges in Massachusetts making sentencing decisions based on “data-driven” analysis.

Earlier this month, Judge John T. Lu — who chairs the Massachusetts Sentencing Commission, reconstituted in 2014 in order to bring a “critical and data-based lens” to the state’s sentencing practices — ordered a criminal defendant who had been convicted of assault and battery to undergo full-scale Ohio Risk Assessment System testing as part of his sentencing process.

ORAS testing involves evaluating a series of factors, including the offender’s criminal history, education, employment, family support, substance use, and neighborhood and/or peer associations. The results are used to provide a statistical measure of the person’s risk of recidivism.

Until Lu’s ruling, ORAS had been used in Massachusetts only to determine conditions for probation.

“Relying on data-based tools is consistent with the court’s historical reliance on empirical approaches, such as its use of the Sentencing Guidelines,” Lu wrote in a memorandum accompanying his order. “Using a risk assessment instrument, in contrast to a subjective clinical evaluation, further provides an objective estimate of risk of re-offense because it is the product of mathematical scores that classify an individual as having either a high, moderate or low risk of reoffending.”

Lu also found that the benefits of ORAS testing outweighed the potential risks in the case before him, but ordered that the defendant be allowed to assert his Fifth Amendment rights by refusing to answer certain questions and to have his attorney present during testing.

Essex County District Attorney Jonathan W. Blodgett said he was “baffled” by the order.

“It’s a tool that the Probation Department uses, and it’s a fairly new tool, so I’m surprised it’s being used in this fashion,” Blodgett said. “I’d like to understand why he’s using it this way and for what purpose.”

Defense counsel Christopher L. Malcolm of Boston declined to comment on the record, citing the ongoing nature of the case.

But William D. Kickham, a criminal defense attorney in Westwood, said he was concerned that applying such statistical tools to sentencing could undermine individualized justice.

“If this becomes a trend, it would cause me significant concern that my client was being treated more as a statistical piece of data rather than as an individual with his or her own history,” he said.

The six-page decision is Commonwealth v. [Redacted], Lawyers Weekly No. 12-034-16. The full text of the ruling can be ordered by clicking here.

Mandatory testing

On April 1, a jury in Essex County Superior Court found the defendant, whose identity has been shielded by the court, guilty on two counts of assault and battery.

That same day, Lu ordered the defendant, pending sentencing, to undergo full-scale ORAS testing.

As the judge outlined in his order, an individual undergoing ORAS testing is presented with seven categories of questions to assess his likelihood of recidivism.

The categories include the offender’s criminal history, education, employment and financial situation, family and social support, neighborhood problems, substance abuse history, peer associations, and criminal attitudes and behavioral patterns.

Within the seven categories lie more than a dozen variables that social scientists have connected to recidivism, including criminal companions, adult criminal history, race, family-rearing practices, age, intellectual functioning, family criminality, gender and socioeconomic status of origin.

At the conclusion of the assessment, the offender receives a score in each category along with an overall assessment classification of either “high,” “moderate” or “low” risk of recidivism.

Acknowledging the ongoing debate over how useful data-driven systems might be in sentencing criminal defendants, Lu stated in his order that ORAS results are not “singularly determinative” of the sentence a court would impose, but serve as one factor to consider.

Lu also acknowledged that data-driven and evidence-based analysis while “not foreign” to the criminal justice system, has traditionally been used for “back end” applications such as guiding parole board decisions, as opposed to “front end” applications such as sentencing decisions.

However, more and more jurisdictions are taking data-driven analyses into account in sentencing, which could potentially make corrections more effective while reducing administrative costs and deterring recidivism, Lu wrote.

“These risk assessments, in theory, yield consistent and reliable results regardless of the individual conducting the assessment,” he continued. “Judicial efficiency is also supported because risk assessment testing provides all judges with identically reported data. Finally, risk assessment helps to reduce the imposition of excessive punishment, which in turn results in reduced incarceration, saved taxpayer monies and reduced damage to individual offenders, their families, and communities.”

At the same time, Lu conceded that the use of tools such as ORAS pose certain risks when incorporated into sentencing.

“By basing sentencing decisions on static factors and immutable characteristics — like the defendant’s socioeconomic background, or neighborhood — they may exacerbate unwarranted and unjust disparities that are already far too common in our criminal justice system and in our society,” Lu said.

Additionally, he noted, constitutional limitations may forbid consideration of traits such as race, sex, religion and socioeconomic status that are the subject of risk-assessment questions.

“Under the circumstances in this case, however, the benefits of having the defendant undergo the ORAS for consideration during his sentencing outweigh the potential consequences,” Lu wrote. “[T]he court is comfortable subjecting the defendant to full-scale testing with safeguards — allowing the defendant to assert Fifth Amendment rights by refusing to answer certain questions and having his attorney present — in place.”

Cause for concern?

Blodgett said he was surprised by the order since ORAS is such a new tool that even probation officers are still being trained on its use.

Accordingly, Blodgett said he wondered if Lu was using ORAS here because he had already determined that the defendant would be sentenced to probation rather than incarceration.

“That’s a concern you’d need to ask about,” he said. “[ORAS] seems to be used here in a different fashion than what it’s intended for.”

Randy Gioia, deputy chief counsel of the Public Defender Division at the Committee for Public Counsel Services, applauded Lu’s desire to reduce the unnecessary punishment of individuals and the damage it causes to their families and communities.

“We need to reduce the reliance on incarceration as a criminal sanction,” he said. “We need to encourage judges to look to alternatives to incarceration that will give people caught up in the criminal justice system meaningful opportunities to obtain good jobs, adequate housing, useful education, and treatment for mental health and substance abuse issues.”

Nonetheless, Gioia expressed concern that a risk assessment tool such as ORAS might not accomplish those goals.

For example, he said, ORAS utilizes race, gender and age — characteristics that may be constitutionally prohibited — alongside other characteristics such as intellectual functioning, family criminality and socioeconomic status of origin — which an individual has no control over — in order to determine a defendant’s risk of recidivism.

Other characteristics the tool utilizes carry an underlying assumption that a person cannot change and is predestined to reoffend, Gioia said.

“I think we should proceed cautiously,” he said. “Let’s make sure this risk assessment tool will accomplish what we want in Massachusetts before judges use it to assist them in making life-altering sentencing decisions.”

Kickham said he worries that a judge might rely on a tool like ORAS to support a more severe sentence than he otherwise would have given.

“To imprison someone because of crimes he’s expected to commit based on statistical inference or statistical studies of general populations denies that person the opportunity to not commit those crimes,” he said.

Meanwhile, Blodgett questioned the utility of the testing given the judge’s emphasis that the defendant could assert his Fifth Amendment privilege and refuse to answer questions.

“If the assessment is such that [the defendant] doesn’t have to answer questions, how much data is actually going to be obtained that would be useful in a situation like this?” he asked. “That’s important. If the whole idea is that we’re using data-driven analysis and he only answers two out of 10 questions, how’s that useful?”

Commonwealth v. [Redacted]

THE ISSUE: Could a criminal defendant convicted of a felony be ordered to undergo full-scale Ohio Risk Assessment System testing as part of his sentencing process?

DECISION: Yes (Superior Court)

LAWYERS: Kelleen Forlizzi of the Essex County District Attorney’s Office (commonwealth)

Christopher Lee Malcolm of Boston (defense)